香港八大 GenAI 政策官方規例對照 2026

香港八大 AI/GenAI 政策同 declaration template 對照——官方規例逐校解讀(2026 更新)

本文由 UniWriterPro 學術規例研究組 verify · Last updated: 2026 年 8 月 16 日 · 本文係按 HKU TALIC + Faculty of Arts GenAI Guidelines、CUHK Academic Honesty Policy、HKUST Academic Regulations、PolyU Student Guide on Academic Integrity + AR GenAI Guidelines、CityU Academic Honesty Rules + GenAI Guidelines、EdUHK Policy on Academic Honesty + LTTC Declaring AI Tools、HKBU CHTL Guidelines + BBA Declaration、HKMU Guidelines for Use of Generative AI(2025 年 9 月生效)、Lingnan University Guidelines for Using GAI Tools 逐條解讀。冇公開文件嘅位一律標「搵唔到官方文件」,唔搬其他學校套用。

好多同學搜「香港大學 AI 政策」得到嘅答案,係一堆抄嚟抄去、冇 verify 過嘅泛化建議——「用 AI 要 declare、Turnitin 超過幾多 % 會 fail、AI detection 好準」。呢啲全部係 唔準嘅網上傳聞。真實情況係:香港八大 每一間 AI/GenAI 政策都唔同——declaration 內容要求、Turnitin AI Score cutoff、subject teacher discretion 空間、prohibited uses 每校都自訂。本文將 8 大官方 GenAI policy verbatim quote 齊,每校 declaration template 逐個對照,讓你按你自己嘅 university 睇你真正要跟嘅規矩。

呢篇會答到你

  • 八大 AI/GenAI 政策整體 approach 有咩分別?「open and forward-looking」vs「course-by-course」vs「declaration mandatory」邊個屬邊間?
  • 每間大學 declaration 具體要寫幾多項?(HKU 6 項 vs EdUHK 4 元素 vs HKBU 兩選一 vs PolyU 三件事)
  • 用 AI 唔 declare 會點?八大明訂罰則有咩共通點?
  • Turnitin AI Score 幾多 % 就一定被指控?(scoop:八大冇一間公開 cutoff)
  • Faculty vs University-wide policy — 我 Faculty 有冇 override 校方?
  • 畢業班/論文班用 AI 嘅特別注意——research integrity 條例點樣同 undergraduate GenAI 政策整合?
  • 搵唔到官方文件嘅位會直接講,唔會編數字。

香港八大 AI/GenAI 政策 quick view 對照表

重點

睇呢張表之前記住:「Declaration 要求」冇一間係 optional——用咗 AI 就一定要 declare。唯一 differ 嘅係 declaration 具體要寫咩由邊個 body 制定Turnitin 使用有冇 mandatory

大學 整體 stance Declaration 具體要求 Declaration submit 方式 Turnitin/AI detector 官方立場
HKU Senate-endorsed;TALIC 訂 core principles;Faculty-specific detail Faculty of Arts:6 項(tool + version、date、full prompt、edit level、affected section、justification) Signed declaration append to submitted work(instructor 可以提供 template) AI detector 只算 supplementary evidence;primary evidence 係 declaration 缺失 + 寫作風格突變 + hallucination + oral defense
CUHK Course-by-course guidance;academic honesty policy 統管 由 course outline/assessment guide 訂;冇校方統一 mandatory template Course-specific;VeriGuide 為主要 similarity check tool VeriGuide 全校用;similarity cutoff 冇公開;AI detection 冇公開 threshold
HKUST 搵唔到官方文件 有 university-wide student-facing rule;faculty resource 唔等於全校規 Course teacher 個別提供;冇校方統一 Course-specific 冇公開全校統一 Turnitin/AI cutoff;submission workflow 以當期 course 文件為準
CityU Academic Honesty Rules 4 條明訂 AI 相關 offence;Effective from Semester A 2025/26 由 Course Leader decision;GenAI Guidelines 提供框架 Course-specific;Assessment brief 明訂 冇公開統一 Turnitin similarity cutoff;AI detection cutoff 搵唔到官方文件
PolyU Open and forward-looking stance;embrace GenAI as positive force in education Subject teacher discretion;三件事必寫:model used、how applied、integration 方式 Assignment submit 時 include;Official declaration template 可 download 冇校方統一 Turnitin cutoff;subject teacher 個別訂;Best practice:即使 assessment brief 冇要求都主動 declare
HKBU University Senate approved;透過 CHTL 頒布;Default 係「Allowed」(explicitly enabled by University) Assignment 內附標準 declaration(兩選一:「did knowingly use」或「did not use」) Assignment 附 signed declaration;BBA 網頁完整版本 CHTL 明文:AI detection tools are NOT to be used as SOLE evidence;需要 oral examination verify
EdUHK Policy on Academic Honesty Section 5.7;用 AI 唔 declare 直接當 plagiarism 4 元素(AI tools 名、purpose、prompts used、integration of outputs) Signed declaration in submitted work + Turnitin Moodle upload 時 select AI usage option 冇公開統一 Turnitin similarity cutoff;AI detection tool(例如 Turnitin AI Score)threshold 冇公開
HKMU Guidelines for Use of Generative AI(Effective 2025 年 9 月 1 日);OLE-centric declaration model OLE 上 submit declaration;如用 AI 要 specify tool + how it contributed(idea generation / concept explanation / structuring / editing) OLE submission 時 declaration——全校 mandatory HKMU AI Writing Detector(combines multiple detection tools)display in OLE;similarity cutoff 冇公開
Lingnan Guidelines for Using GAI Tools + Best Practices(TLC 發布);Course syllabus mandatory 提及 由 course syllabus specify;冇校方統一 mandatory template Course-specific;配合 Turnitin submit Turnitin 為 writing assignment mandatory tool;similarity cutoff 冇公開;AI detection cutoff 冇公開

睇完呢個 overview,你會發現八大共通嘅係「declaration 一定要」+「course teacher 有 discretion」;差異最大係 declaration 內容具體要求——由 HKU 6 項到 HKBU 一句話兩選一,差別實在好大。以下逐間拆開講。

HKU:Senate-endorsed;Faculty of Arts 6 項 declaration 最詳

HKU 兩層架構

Layer 1:TALIC(Teaching and Learning Innovation Centre)Core Principles——University-wide statement。Layer 2:Faculty-specific guidelines——Faculty of Arts 2025 版係全港最詳細嘅 declaration template。

Staff and students are reminded that the HKU regulations and policies regarding plagiarism and academic mis-conduct supersede all other regulations, where there is a conflict, these regulations and policies take precedence.

HKU TALIC — AI Ethics Core Principles

ANY misuse of AI or misrepresentation of AI work as one’s own independent creative work violates the HKU regulations as well as generally accepted principles of community, integrity and honor. Violations will be referred to the university disciplinary committee.

HKU TALIC — Penalty and Referral

Faculty of Arts 2025 版 6 項 mandatory declaration——係全港 undergraduate AI declaration template 之中最詳細:

Whenever GenAI contributes in any way to submitted work, students must append a completed GenAI Use Declaration (a template may be offered by instructors). The declaration must be signed and include, at minimum: 1. Tool(s) and version used (e.g., ChatGPT-4o, Gemini 1.5 Pro). 2. Date(s) of interaction. 3. Full text of the exact prompt(s) that generated material subsequently used. 4. Description of what was kept verbatim, heavily edited, lightly edited, or discarded. 5. Precise indication of which sections or elements of the submission (by paragraph number, line number, or appendix) received AI contribution. 6. Brief justification of why GenAI was used at each stage.

HKU Faculty of Arts — GenAI Declaration 6 Requirements

呢一句最有價值——因為佢明確話 AI detector 唔係 primary evidence

The Faculty does not exclusively rely on commercial “AI detectors” because of their high false-positive and false-negative rates. Turnitin’s AI-writing report may be consulted as supplementary evidence only. Cases will be evaluated primarily on: Absence or inadequacy of the required GenAI Use Declaration. Sudden, unexplained shifts in writing style, vocabulary, or sophistication. Presence of typical GenAI errors (factual hallucinations, repetitive phrasing, anachronistic references). Inability of the student to explain or defend their submitted work orally or in writing when requested.

HKU Faculty of Arts — Evidence Basis for AI Misconduct

對 HKU Arts 同學嘅實務意義:Turnitin AI Score 高並唔等於 automatic misconduct——你 declaration 內容齊、寫作 consistent、可以 oral defense、冇 fabricated citation,就算 AI Score 60% 都可以 defend。反之 declaration 缺失就算 AI Score 5% 都可以被查。想深入了解 HKU 全部 AI 政策細節,請睇 港大 HKU GPA、CGPA、GGPA、Honours 5 級分類、AI 政策官方規例解讀

CUHK:Course-by-course discretion;academic honesty 統管

CUHK 冇一份 University-wide 學生版 GenAI declaration mandatory template——由 course outline 同 assessment guide 訂具體做法。VeriGuide 係全校 similarity check tool,但係 cutoff 冇公開。

對 CUHK 同學實務 takeaway:每份 assessment 前對返 course outline——subject teacher 個別可能訂:「呢份 assignment 不允許 GenAI」、「允許但要 declare」、「允許用作 brainstorming 但唔可以 submit AI-generated content」——三種完全唔同 approach 可能出現喺同一 department 唔同 course。詳情請睇 中大(CUHK)GPA、重讀、補考、AI 政策官方規例解讀

搵唔到官方文件 全校統一「VeriGuide 超過 X% 必 fail」數字;搵唔到官方文件 一個 University-wide student-facing mandatory GenAI declaration template(不同 Faculty/Department 自訂);搵唔到官方文件 AI-detection tool 全校 cutoff。呢啲位以 course outline 為準。

HKUST:University-wide student-facing rule 官方留白

科大情況特別——就 學生用 AI/Generative AI 統一政策,公開 official document 冇一份 University-wide student-facing rule。faculty-facing resource 例如 Center for Education Innovation 頁面針對 teacher,唔應該當成全校學生規例。呢個 gap 對學生實務上意味住:你必須逐個 course 睇 course outline,唔可以假設一 course allow 就其他都 allow。

搵唔到官方文件:科大 university-wide student-facing GenAI mandatory declaration template;全校 Turnitin similarity cutoff;AI-generated content detection cutoff。呢三項都要以當期 course 文件為準。詳細請睇 科大 HKUST CGA、重讀、AI 政策官方規例解讀

科大同學想 outsource 學術寫作嘅實用 note:科大 submission workflow 完全 depends on course——即係如果你 course outline 冇 mandatory declaration 要求,你交嘅 essay/FYP report 只需要 meet 該 course 個 rubric。想拎專業 human-writer 起草嘅 essay/case study/FYP,可以睇 Essay 代寫Coursework 代做

CityU:AR 4 條明訂 offence + Effective Sem A 2025/26

城大 Academic Honesty Rules 明文列 AI 相關 offence 有兩條特別要記住:

Using materials generated by artificial intelligence (AI) technology, including but not limited to AI-generated fabricated citations and references, as one’s own work, or presenting AI-generated result that has not been expressly permitted by the concerned academic staff as one’s own work.

CityU Academic Honesty Rules — AI offence(Effective Sem A 2025/26)

兩個 element 特別注意——,AI-generated fabricated citations(AI 虛構嘅 reference)都明文列做 offence——同 HKU Faculty of Arts approach 相同;,「expressly permitted by the concerned academic staff」——即係話 course leader 明文 permit 咗至可以用,冇 explicit permit 就當 offence。

Effective Sem A 2025/26 呢個 date 好重要——2025 年 8 月 31 日之前入學嘅同學,某啲 legacy 課可能仲用緊舊版 Academic Honesty Rules;由 Semester A 2025/26 起所有 course 都套用新版 AI offence 條文。詳情請睇 城大 CityU GPA、Advanced Standing、Honours、AI 政策官方規例解讀

PolyU:Open and forward-looking stance;subject teacher discretion

理大 GenAI 政策採取 open and forward-looking 立場——校方主動 embrace GenAI 作為 educational tool,但同時強調三個核心原則:students accountable for their own work、properly acknowledged when used、subject teacher discretion to restrict

PolyU takes an open and forward-looking stance on the use of generative artificial intelligence (GenAI) tools (such as ChatGPT) as a positive and creative force in education, and the incorporation of such use in innovative learning, teaching, and assessment practices. While embracing the use of new technology in education, PolyU upholds the principle that students must adhere to high standards of academic integrity in all forms of assessments. The Student Code of Conduct and the policy on academic integrity apply to the use of GenAI in student work. In accordance with the above stance, the University emphasises that: students are accountable for their own work; and the use of GenAI (if permitted) in student work should be properly acknowledged.

PolyU Student Guide on Academic Integrity — GenAI Policy

PolyU 2025 年 2 月 update 咗涵蓋 AI translation tools——即係話用 AI 做翻譯(例如將中文論文 translate 做英文)都要 declare:

PolyU maintains its open and forward-looking stance on the use of GenAI in education. This update highlights the principle that the work students submit for assessment must be their own work and that any AI-generated material (including text, images, and translations) used in the work must be properly referenced. Subject teachers have the discretion to restrict the use of AI tools and/or AI-generated material. If the use of GenAI should be restricted, subject teachers will clearly communicate the restrictions to their students in the assignment briefs.

PolyU AR — Guidelines for Students on GenAI (2025 Updates)

PolyU declaration 具體 template 冇 mandate 全校統一,但 declaration 內容有三件 mandatory 事,GenAI model used(e.g. ChatGPT-4、Claude、Gemini);,how it was applied(brainstorming、drafting、editing、translation);,integration 方式(direct paste、paraphrase、reference)。詳情請睇 理大 PolyU GPA、Weighted GPA、Award GPA、Honours 分類、AI 政策官方規例解讀

HKBU:Default 係「Allowed」;兩選一 declaration template

浸大 AI 政策特殊之處:default position 係 allowed——即係話除非 assignment brief 明文禁,否則你可以用。呢個同其他大學「default restricted, permit 至可用」剛好相反。

In alignment with the University’s guiding principles, the default approach is that students are ALLOWED and empowered to use generative AI tools in assessed assignment tasks.

HKBU BBA — CHTL Generative AI Guidelines

HKBU declaration 係兩選一 template——即係話用咗 or 冇用都要 declare:

I did knowingly use generative AI tools in this assignment. I followed the University’s guidelines for students on academic integrity. No content generated by generative AI tools has been presented as my own work. I take responsibility for the work submitted.

HKBU BBA — Declaration Template (Used AI)

OR I did not use generative AI tools in this assignment.

HKBU BBA — Declaration Template (Did Not Use AI)

浸大 AI detection 政策亦係全港最 explicit 對學生有利嘅一句——直接寫明 AI detection tools 唔可以做 SOLE evidence

AI detection tools are NOT to be used as SOLE evidence of improper use of generative AI tools. Other methods, such as oral examination, must be used to verify cases of improper use.

HKBU CHTL Guidelines — AI Detection Cannot Be Sole Evidence

對浸大同學實務意義:如果你被 Turnitin AI Score flag,佢必須安排 oral examination 或者其他 verify 方法先可以 finalize misconduct decision——呢句係浸大同學 hearing 時最有力嘅一個 defense point。詳情請睇 浸大 HKBU GPA、重讀、補考、AI 政策:官方規例逐條解讀

EdUHK:4 元素 declaration + Turnitin OLE upload 時 select

教大係香港八大 declaration format 之中最結構化嘅——4 元素 mandatory + Turnitin Moodle upload option 兩層設計:

Use of AI-Enabled Generative Tools without Proper Declaration: If you use AI-Enabled Generative tools in the preparation of an assessment task, you must declare your use of such tools and take full responsibility for the content submitted for assessment. The acceptable use of AI varies across disciplines, subjects, and assessment tasks. Course teacher will provide this information. But it is your responsibility to check the assessment guidelines and relevant policies, and to understand what is expected of students.

EdUHK Student Handbook — Academic Honesty and Copyright

Copying from AI-generated contents without proper declaration is also regarded as plagiarism.

EdUHK Student Handbook — AI Copy Without Declaration = Plagiarism

EdUHK LTTC 具體 declaration template 明訂 4 元素:

Students are required to provide a clear declaration when AI tools have been used in the preparation of an assessment. This declaration should be included in their submitted work. To ensure transparency, it is recommended that the declaration should include the following elements: AI Tools: Specify which AI tools were employed in the assessment. Purpose: Describe the intended use of these AI tools in the assessment. Prompts Used: Detail the prompts entered into the AI tools. Integration of Outputs: Explain how the outputs generated by the AI tools were integrated into the submitted work.

EdUHK LTTC — Declaring Use of AI Tools (for Students)

Turnitin Moodle upload 時亦要 select AI usage option:

Students are required to acknowledge if their Turnitin assignments contain materials generated by AI tools. This can be made by selecting the appropriate option on Moodle during the assignment upload process.

EdUHK LTTC — Turnitin Moodle Upload AI Option

詳情請睇 教大 EdUHK GPA、Honours、Field Experience、AI 政策官方規例解讀

HKMU:Guidelines for Use of GenAI(2025 年 9 月生效);OLE-centric

都會大學 GenAI 政策 Effective Date 係 2025 年 9 月 1 日——比其他大學晚 endorse,但係結構清晰。5 項 mandatory principles 全部歸類為學生責任:

Uphold Academic Integrity: Students must always act with honesty and integrity in all academic work, strictly complying with the University’s Regulations Governing Academic Integrity and Student Discipline. Any inappropriate use of AI, including using AI without proper acknowledgment or in violation of course guidelines, will be treated as academic misconduct.

HKMU Guidelines for the Use of Generative AI (Students) — Section 2.1

Declare and Acknowledge AI Use: When completing assignments, students are required to submit a declaration on the Online Learning Environment (OLE) indicating whether generative AI tools were used. If AI was utilised, students must specify which tool was used and describe how it contributed to their work—such as generating ideas, explaining concepts, structuring assignments, or editing content. This transparency ensures responsible use and allows teaching staff to assess the originality and authenticity of student work.

HKMU Guidelines for Use of GenAI — Section 2.2 OLE Declaration

Adhere to Course-Specific AI Guidelines: Students must strictly follow the approach to AI use outlined in their course materials. This may include prohibiting AI use entirely, permitting it only under certain conditions, or requiring explicit acknowledgment.

HKMU Guidelines for Use of GenAI — Section 2.5 Course-Specific

HKMU 特殊之處:OLE 已內建 AI Writing Detector——combines multiple detection tools,結果 display 喺 OLE 內:

AI Writing Detector: Combines multiple detection tools to assess the likelihood that student work was generated by AI. Results are displayed within the OLE, supporting academic integrity and efficient evaluation.

HKMU AI Tools for Teaching Staff — AI Writing Detector

對 HKMU 同學嘅實務意義:OLE 上 submit 每份 assignment 都需要 declaration checkbox——即使冇用 AI 都要 tick「沒有使用」。搵唔到官方文件 有明確 AI Writing Detector cutoff percentage;搵唔到官方文件 Turnitin similarity 全校 cutoff。ChatGPT-4o 已 through HKMU ChatGPT Web Portal 提供畀學生 free access——即係校方主動 embrace GenAI 但要求 responsible use。

Lingnan University:GAI Guidelines + Best Practices(TLC 發布)

嶺南大學 AI 政策一大特色:2023 年 7 月起免費提供 ChatGPT 畀所有 staff + students——包括 GPT-4.0、GPT-4.0 Turbo。同時 course syllabus 必須明文提及 AI 使用規例:

Since July 2023, all staff and students at Lingnan University have been given free access to ChatGPT, followed by GPT 4.0 and GPT 4.0 Turbo. Lingnan University has introduced several AI-related disciplines, as well as mandatory generative AI courses for all freshmen.

Lingnan Press Release 2024 — Free ChatGPT + Mandatory GenAI Course

Lingnan course syllabus template 統一寫明:

Students must adhere to the University’s guidelines and practices when using Generative Artificial Intelligence (AI) tools. The official documents “Guidelines for Using GAI Tools at Lingnan University” and “Best Practices for Ethical and Responsible Use of GAI Tools in Course Assessments” can be found here.

Lingnan Syllabus Tool — Standard AI Clause

對 Lingnan 同學嘅實務意義:每份 course syllabus 都會有一段 AI 條款——即係話 subject leader 有 discretion 訂具體規例,同時所有 writing assignment 都要 through Turnitin submit。搵唔到官方文件:Lingnan University-wide student-facing mandatory declaration format;搵唔到官方文件:Turnitin similarity cutoff 或 AI-detection tool threshold——由 subject leader 個別訂。

八大 declaration template 具體對照

以下呢張 table 係整篇最實用嘅位——你可以直接對照自己大學嘅要求,睇你 declaration 要寫嘅具體 field:

大學/Faculty Declaration 必寫 field Submit 方式 Prohibited use(如明列)
HKU Arts Tool + version、date、full prompt、edit level、affected section、justification(6 項) Signed declaration append to submission AI 生成 references/bibliographies;misrepresenting AI work as own creation
CUHK 由 course outline 訂——冇校方統一 mandate Course-specific;配合 VeriGuide 由 course outline 訂
HKUST 由 course outline 訂——冇校方統一 mandate Course-specific 由 course outline 訂
CityU 由 Course Leader 訂;AR 明文列 AI-generated fabricated citation = offence Course-specific;Effective Sem A 2025/26 AI-generated result 未 expressly permitted;AI 虛構 citation
PolyU GenAI model + how applied + integration 方式(3 項) Assignment 附 declaration;Official template download available 用 AI 做整份 assignment;直接複製 AI-generated content
HKBU 兩選一(使用 / 冇使用);simple template Assignment 附 signed declaration;BBA 網頁完整版本 用 AI 但唔 declare;presenting AI content as own work
EdUHK AI tools 名 + purpose + prompts used + integration of outputs(4 元素) Signed declaration + Turnitin Moodle option(雙重) 用 AI 但 improper declaration = plagiarism
HKMU OLE checkbox + tool + contribution 描述(idea generation/structuring/editing) OLE 上 submit declaration——全校 mandatory 用 AI 但冇 OLE declaration;違反 course-specific 限制
Lingnan 由 course syllabus specify Course-specific;Turnitin mandatory for writing assignment 由 course syllabus 訂

八大同學最常犯嘅 5 個 AI declaration 錯誤

錯誤一:以為 declaration 寫「用咗 ChatGPT」就夠。絕大部分大學(HKU、EdUHK、PolyU)都要具體 field——tool name、version、prompt、integration 方式。HKU Arts 更加要 6 項。一句「有用 ChatGPT」係 inadequate declaration,可以直接 trigger investigation。

錯誤二:以為 Turnitin AI Score 低就唔洗 declare。錯——AI Score 唔係決定 declaration 是否需要。用咗 AI(無論 how much)就要 declare。校方唔係 chase 你 AI Score 高,係 chase 你 declaration 是否 truthful and complete

錯誤三:以為冇 declare 都可以 defend「唔知有呢 policy」。八大全部明文將政策放喺 Student Handbook 或 Academic Registry 網頁——「唔知」唔係 defense。HKU/EdUHK/HKMU 甚至將 undisclosed use 直接 equate 到 plagiarism。

錯誤四:以為 course outline 冇提就等於 AI 用得。唔對——HKU、EdUHK、CityU 都係 default restricted, permit 至可用 approach——即係 course outline 冇 explicit permit 就當唔可以用(HKBU 例外,係 default allowed)。

錯誤五:用 AI 生成 citation/reference。八大之中至少 3 間(HKU Arts、CityU、CUHK)明文將 AI 虛構 reference 列做 misconduct offence——因為 ChatGPT 經常 hallucinate 唔存在嘅 source。你 draft 用 AI 補 citation,一 verify 就 collapse。

跨校 pattern:3 個 axis 睇八大差異

將八大 AI 政策 collapse 成三條 axis,你就 map 到自己校情況:Axis 1(default 態度)——HKU/EdUHK/CityU/HKMU 屬「default restricted, permit 至可用」;HKBU 屬「default allowed,除非 course outline explicit ban」;CUHK/HKUST/PolyU 屬「Course-by-course discretion」冇 university-wide default。Axis 2(declaration 具體度)——HKU Arts 6 項最詳、EdUHK 4 元素、HKBU 兩選一 template;其餘 course-by-course。Axis 3(Detection tool)——HKMU 有自研 AI Writing Detector;CityU/HKU 明文提 Turnitin AI;其餘冇公開統一 threshold。你入學前睇一次自己 course outline,行 3 條 axis 對照,就知自己屬邊 type、要唔要主動 declare。

UniWriterPro 點樣 fit 入你 declaration workflow?

學術寫作幫手同代做代寫係唔同性質嘅嘢——UniWriterPro 唔會叫你去大學 declaration form 上寫「用 AI generate 咗全文」;我哋提供嘅係 human-writer 起草嘅代做代寫服務。你收到嘅稿係按你 assignment brief 由 human writer 起草,符合 Faculty citation format 同 methodology requirement——剩落嘅 declaration 決定係你自己嘅 personal choice。

Proofreading 特別安全——因為 human editor 改文法錯字唔屬於 AI-generated content。呢個係最保守嘅使用方式,適合 non-native English speaker 交 essay/FYP/thesis final draft 之前做語言 polish。可以參考 Proofreading & 修改 服務。

其他常用服務:Essay 代寫(3000+ words assignment)、Coursework & Case Study 代做(case analysis、Group project 個人部分)、FYP 全程支援(Business、Engineering、Applied Physics、Nursing)、Thesis 論文代寫(Master’s dissertation、Doctoral thesis)、Presentation & Speech 代做(含 PPT slides 同 speech script)。全部由 human writer 起草,唔用 AI generate。

其他官方冇明講嘅位(八大共通)

以下呢啲位八大都搵唔到官方公開文件,唔係「唔存在」,係「冇公開統一 threshold」。呢種情況唔會自動 penalty,但一旦有 investigation 你就要自己提供解釋同證據。呢啲位如果你想確認,直接寫 email 去對應 office(Academic Registry、Programme Leader、subject leader),比睇 forum/reddit/小紅書可靠,亦都可以保留書面 written record 以備日後:

  • Turnitin similarity cutoff——八大冇一間公開「超過 X% 自動 flag」全校數字;由 subject leader/course teacher 個別判斷
  • AI-detection tool(Turnitin AI Score、GPTZero、Turnitin AI Writing Score)cutoff——八大冇一間公開全校 threshold
  • Prohibited AI tool specific list——冇一間大學公開「呢個 AI tool 唔可以用」list(除咗個別 course level restriction)
  • Declaration late submission grace period——冇校方統一 grace period;即係話 declaration 冇 submit 都可能被查
  • Retrospective policy application——如果你 2023 年入學但 2025 年畢業,你要跟邊個 version?八大冇明訂 grandfather clause

共通點:各 course/subject/programme 有自主權,冇 University-wide 統一 threshold。想追根究底就查你自己 Course Outline/Subject Syllabus/Programme Handbook 對應章節,或者直接 email 對應 office 拎書面確認。

實操建議:你入學前後行三步 preparation——第一步,登入自己 programme 官方 handbook,Ctrl+F 搜「AI」「GenAI」「generative」「ChatGPT」「declaration」五個 keyword,睇睇你 programme 有冇補充規則。第二步,逐科 course outline 睇「AI use policy」/「Use of GenAI」/「Academic Integrity」呢三個 section——你都係上到 Semester Week 1 見到該 subject teacher 具體 stance,就已經 default 到最新官方規則。第三步,將對應規則 URL 保存到自己 note app(Notion/Apple Notes/Evernote),交功課寫 declaration 時就直接拉出來對照 template,唔會 last minute panic。呢個 workflow 完全 zero-cost,但可以救你成個 learning cycle 嘅風險。

FAQ — 常見問題

Q1. 用 AI 但 declaration 只寫「有用 ChatGPT」夠嗎?

絕大部分大學係 唔夠。HKU Arts 要 6 項、EdUHK 要 4 元素、PolyU 要 3 件事。每個 field 都要具體:tool name(e.g. ChatGPT-4o、Claude 3.5、Gemini Advanced)、version(date-based version 更好)、how it was applied(brainstorming/drafting/editing/translation)、integration(direct paste/paraphrase/reference)。如果你 declaration 淨係寫「有用 ChatGPT」,喺 HKU/EdUHK/PolyU 就算 declaration inadequate,可以直接 trigger investigation。

Q2. Turnitin AI Score 高過幾多 % 一定被指控?

八大都冇一間公開 cutoff 數字——即係話冇一個 X% 係 automatic misconduct trigger。HKU Faculty of Arts 明文話 AI detector 只算 supplementary evidence——primary evidence 係 declaration 缺失、寫作風格突變、hallucination、oral defense failure。HKBU 更加明文話 AI detection tools NOT to be used as SOLE evidence——即係話 AI Score 高但 declaration 齊、oral 可以 defend,都可能唔會 misconduct finding。反之:declaration 缺失就算 AI Score 5% 都可以被查。

Q3. 用 AI 生成 idea 然後自己重寫,仲需要 declare 嗎?

Yes,八大都話「in any way」contribute 都要 declare。HKU Faculty of Arts 明訂 declaration 要 describe「what was kept verbatim, heavily edited, lightly edited, or discarded」——即係即使 heavily edited 過都要 declare。EdUHK 亦要 describe「how the outputs were integrated」。安全 approach:只要 AI 有 touch 過你 draft(brainstorming、outline、grammar check、translation),就 declare;冇罰則 for over-declaration,反而保護你唔會被 accused as undisclosed use。

Q4. 我 Faculty 冇 explicit GenAI policy,可以按 University-wide policy 為準嗎?

可以,但要 combine 三層:University-wide statement(例如 HKU TALIC)+ Course outline explicit clause + Assessment brief specific requirement。如果三者衝突,一般 order of precedence 係:Assessment brief > Course outline > University-wide policy——即係 subject leader 落 assessment brief 個 rule takes priority。呢個係 course-by-course discretion 嘅本質。

Q5. 用 AI 做 translation(e.g. 中譯英)需要 declare 嗎?

Yes——PolyU 2025 年 2 月 update 明文將 AI translation 列為 AI-generated material,需要 properly reference。HKU Faculty of Arts 亦明訂「AI 生成 text, ideas, structure, translations, or code」都要 declare。用 Google Translate 或 DeepL 呢類「classic translation tool」歷史上唔算 AI-generated(但呢個判定 evolving);用 ChatGPT/Claude/Gemini 做 translation 一定要 declare。

Q6. 用 AI 補 citation/reference 有咩風險?

最危險位——HKU Faculty of Arts、CityU AR、CUHK academic honesty policy 都明文將「AI 生成 fabricated citation」列為 misconduct offence。ChatGPT 經常 hallucinate 出唔存在嘅 author name、journal name、DOI——一 verify 就 collapse。任何時候用 AI 補 citation 之後,你必須逐條 verify:Google Scholar 搜到 exact match、library database 見到 real PDF。如果 verify 唔到就一定唔可以 cite。

Q7. Proofreading service 需要 declare 嗎?

Human editor 改文法錯字唔屬於 AI-generated content——冇一間大學要求 declare human proofreader(例如 friend、tutor、professional editor)。但如果你 proofreader 用 AI-powered tool(例如 Grammarly Premium AI、ChatGPT)改文法,就要 declare AI usage。UniWriterPro 嘅 Proofreading 服務 由 human editor 執行,即係 safest option。

Q8. 冇用 AI 但被 flag,點做 defense?

三個 defense evidence,你 writing process document(Google Docs edit history、Word document version、outline draft)——可以 timestamp 你手動打嘅 progression;,oral defense capacity——係咪可以 explain 你論點、terminology 由來、reference 選擇 rationale(HKU/HKBU 明文將 oral defense 列為 primary evidence);,writing style consistency——搵你之前 assignment 對比,如果寫作 pattern 一致就係強 defense。八大之中 HKBU 最有利:CHTL 明文話 AI detector 不能做 sole evidence。

Q9. Group project 每個組員都要交 declaration 嗎?

Yes——每個組員個人。因為 declaration 係 personal responsibility,即係話你自己嘅 contribution 有冇用 AI、你 individually 是否 use,需要 individually declare。八大 declaration 條款都用「student」(individual)而唔係「group」——即係你組員用咗 AI 但冇 declare,佢個人被查;你個人冇用又 truthfully declare 冇使用,你 defend 到。實務上,Group project 交前 group meeting 對齊每個 member declaration statement。

Q10. GenAI 政策未來 6 個月會有 major update 嗎?

八大 GenAI 政策仍係 rapidly evolving space——CityU 2025/26 新版 AR 明文列 AI-generated fabricated citation offence、HKMU 2025 年 9 月生效嘅 Guidelines、PolyU 2025 年 2 月 update translation clause,全部反映 policy 每 6-12 個月都可能有 major amendment。Best practice:每學期 start 前,登入你大學 Academic Registry 網頁 check Academic Honesty Rules 有冇新 version、Student Handbook 有冇 update、你 course outline 有冇明訂 GenAI clause。三個都 review 齊,先按最新 version 準備 assignment。

睇完仲係唔知你 declaration 應該點寫?

你已花咗 15 分鐘睇完八大 declaration template——但你手上有份 essay/FYP/thesis final draft,仲差 3 日就 deadline。與其花另外 5 個鐘 struggle declaration 具體寫幾多,不如直接找 human writer 起草——你收到嘅稿係按你 assignment brief 由 human 寫,符合 Faculty citation format 同 methodology 要求。declaration 決定係你自己嘅 personal choice——我哋唔會 misrepresent 你 course policy。

WhatsApp 5 分鐘拎報價 · 24 小時急件收 · 涵蓋八大 all Faculty · essay/FYP/thesis/coursework/presentation/proofreading 6 大服務

6 大服務:EssayCourseworkFYPThesisPresentationProofreading

Scroll to Top